Security audit
Invariance Battery
Security checks across malware telemetry and agentic risk
Overview
This skill is a small documentation-only guide for checking AI-agent safety invariants and does not install code, request credentials, or run commands.
This appears safe to install as an instructional skill. If you later adapt its audit-trail or action-blocking ideas into actual automation, review where logs are stored and make sure users can understand and override any blocking behavior.
SkillSpector
By NVIDIA
Vulnerability Patterns
- Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
- Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
- Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
- Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
- Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
VirusTotal
63/63 vendors flagged this skill as clean.
Static analysis
No suspicious patterns detected.